Dysautonomia in hypermobile Ehlers-Danlos syndrome and hypermobility spectrum disorders is associated with exercise intolerance and cardiac atrophy

Abstract

Dysautonomia is a recognized manifestation in patients with joint hypermobility (JH) disorders. Symptoms can be highly debilitating and commonly include physical deconditioning and poor aerobic fitness. In this study, the prevalence of dysautonomia, range of associated symptoms, patient-reported physical activity levels and echocardiographic features were assessed retrospectively in a cohort of 144 patients (94% female) with hypermobile Ehlers-Danlos syndrome (hEDS) or hypermobility spectrum disorder (HSD). Echocardiographic parameters of left ventricular size and function were compared between patients with and without dysautonomia, as well as to reported values from healthy controls. Dysautonomia was identified in 65% of female and 44% of male subjects and was associated with a high burden of symptomatology, most commonly exercise intolerance (78%). Exercise capacity was limited by dysautonomia, often postural symptoms, in half of all patients. We observed a reduction in physical activity following the onset or significant flare of hEDS/HSD, most strikingly noting the proportion of dysautonomic patients with sedentary lifestyle, which increased from 44% to 85%. JH-related dysautonomia was associated with smaller cardiac chamber sizes, consistent with previous reports in positional orthostatic tachycardia syndrome. Dysautonomia is prevalent in patients with hEDS/HSD, exercise intolerance is a key feature and leads to drastic decline in physical activity. Unfavorable cardiac geometry may underlie dysautonomia symptoms and may be due to cardiac atrophy in the setting of aerobic deconditioning.

Introduction

Systemic symptoms are common in patients with joint hypermobility (JH)-related disorders including hypermobile Ehlers-Danlos syndrome (hEDS) and hypermobility spectrum disorders (HSDs) (Castori et al., 2017) and include palpitations, atypical chest discomfort, fatigue, dizziness/lightheadedness, syncope, exercise intolerance, dependent acrocyanosis, peripheral vasoconstriction, and impaired thermoregulation (A. Hakim et al., 2017; A. J. Hakim, Cherkas, Grahame, Spector, & MacGregor, 2004). These symptoms are commonly linked to orthostatic intolerance, orthostatic hypotension, positional orthostatic tachycardia syndrome (POTS) or neurally-mediated hypotension (NMH), (A. Hakim et al., 2017) a collection of diagnoses falling broadly under the umbrella of dysautonomia (Robertson, 1999). Dysautonomia symptoms have been reported at high rates (up to 78%) in JH and EDS cohorts (Gazit, Nahir, Grahame, & Jacob, 2003; A. Hakim et al., 2017), with POTS as the most prevalent autonomic profile (De Wandele et al., 2014) (Rowe et al., 1999). Various objective testing modalities are available for the specific subtypes of dysautonomia (i.e. tilt table testing for POTS) but do not correlate with symptom burden (Boris, Huang, & Bernadzikowski, 2020). Therefore, more broadly, dysautonomia is a clinical diagnosis of exclusion based on a classic collection of symptoms and characteristic exacerbation or improvement under expected conditions. Proposed mechanisms include sympathetic neurogenic dysfunction, vascular connective tissue laxity, sudomotor impairment and use of vasoactive medications (De Wandele et al., 2014; Wallman, Weinberg, & Hohler, 2014). One major challenge in the study of dysautonomia is the absence of universal objective diagnostic criteria or biomarkers.

Physical deconditioning and poor aerobic fitness are commonly cited in patients with JH-related disorders, often with dysautonomia symptom onset following a prolonged period of reduced physical activity (e.g., following joint injury)(A. Hakim et al., 2017). In POTS, deconditioning has been linked to small cardiac size with reduced blood volume (Fu et al., 2010) and aerobic reconditioning therapy through exercise can improve or resolve symptoms (George et al., 2016; Winker et al., 2005) while also increasing cardiac mass and blood volumes (Fu et al., 2010; Fu et al., 2011). It is unknown whether small heart size explains the broad dysautonomia accompanying JH-related disorders, or whether aerobic training can offer a similar benefit in this population. Here, we report the prevalence of dysautonomia in hEDS and HSDs according to contemporary nosology, the comprehensive range of associated symptomatology, physical activity levels before and after the onset of hEDS/HSD and cardiac morphologic profiles that may be contributory.

Methods

Included subjects were those seen in a multidisciplinary Cardiovascular Genetics Program for evaluation of a JH-related disorder between January 2017 and January 2020, referred by cardiologists, primary care providers, rheumatologists, orthopedists, pain management specialists, neurologists or self-referred and evaluated by a study investigator (A.R.K, B.D.G. or L.M.). The study was conducted under the approval of Mount Sinai’s Institutional Review Board (GCO # 19–0883 ISMMS). Electronic health records (EHR) were retrospectively reviewed for demographic data and diagnoses of hEDS or HSD documented by a study investigator within the progress note linked to the in-person clinical evaluation. The diagnosis of hEDS or HSD was determined according to the 2017 consensus guidelines including degree of hypermobility based on the Beighton score (“Criterion 1”), presence of systemic and echocardiographic features and family history (“Criterion 2”) and exclusion of alternative genetic or rheutmatologic diagnoses (“Criterion 3”).¹ Ascertainment of clinical dysautonomia was determined through review of EHR documentation (within the progress note or “problem list”) for diagnoses of POTS (including hyperadrenergic POTS), orthostatic intolerance, orthostatic hypotension, NMH or dysautonomia,. Patient-reported symptoms previously described in association with dysautonomia were systematically collected, including: (a) resting sinus tachycardia or palpitations; (b) persistent fatigue; (c) chest discomfort at rest and/or exertion; (d) hypotension from historical data provided by patient or objective data within the her; (e) dyspnea; (f) near or true syncope; (g) dizziness related to postural changes reported with onset by changing from a supine to upright position or with maintaining upright posture; (h) cognitive complaints including concentration impairment (most frequent), “brain fog,” word finding difficulties and/or impaired memory; (i) lower extremity edema and/or discoloration often described as “blood pooling” after standing for relatively short periods of time; (j) temperature intolerance or dysregulation either explicitly reported by patients or as complaints of cold, dusky extremities and/or heat as a trigger for other dysautonomia-related symptoms such as near-syncope; (k) gastrointestinal symptoms including chronic or episodic constipation, nausea and/or diarrhea, (l) complaints of underactive and/or overactive bladder symptoms; (m) exercise intolerance. For this study, we defined individuals with dysautonomia as those having a clinical label of dysautonomia plus reporting symptoms in ≥ 3 of the above thirteen categories. Next, we systematically extracted physical activities reported before the onset or flare of hEDS/HSD symptoms (“baseline”) and at time of the initial evaluation (“current”). Patient exercise levels were categorized as either: 1) none, 2) light, 3) moderate or 4) vigorous based on highest intensity activity reported and as per established guidelines and recommendations (Levine et al., 2015; Maron et al., 2004; Metkus, Baughman, & Thompson, 2010) (Supplemental Table 1).

2-D echocardiographic data was collected from clinical reports where available, to record the following: left ventricular (LV) ejection fraction (EF), LV internal diastolic diameter (LVIDd), LV end-diastolic volume (LVEDV) indexed to body surface area (LVEDV/BSA), posterior wall thickness (PWT) and interventricular septal wall thickness (IVSWT). LV mass was calculated by linear method as recommended by the American Society of Echocardiography (ASE) with the Devereux formula: LV mass = 0.8 × {1.04 [(LVIDd + PWT + IVSWT)³ − (LVIDd)³]} + 0.6, with all measurements obtained in diastole (Lang et al., 2015) and indexed to BSA (LV mass/BSA). Means of each parameter were compared for hEDS/HSD patients with vs. without dysautonomia. Mean LVIDd and LV mass/BSA were compared with previously published gender-specific mean reference normal values derived from non-hypertensive heart-healthy controls (Cuspidi et al., 2012).

Descriptive statistics for categorical variables were reported as frequency and percentages, continuous variables were reported as median and range or mean and standard deviation. Statistical analysis was performed using SPSS version 25 (IBM Corp., Armonk, NY) and Microsoft Excel. Fisher’s exact test (two-tailed) was used to compare rates of dysautonomia between hEDS and HSD cohorts. The Student’s t-test (2-tailed, heteroscedastic or homoscedastic after calculation of variance using the F-test) was used to compare the distribution of echocardiographic values in hEDS/HSD subjects with vs. without dysautonomia. The one sample t-test was used to compare calculated mean echocardiographic values for the cohorts against published reference normal mean values. P<0.05 was considered as statistically significant.

Results

During the two-year study period, 144 patients were evaluated for and diagnosed with a JH-related disorder, 60 (42%) with hEDS and 84 (58%) with a HSD (Table 1). The median age was 31 years (interquartile range 25–42) and 94% were female. In total, 101 (96 female, 5 male) patients (70%) were clinically labeled with dysautonomia, more commonly in hEDS vs. HSD (48, 80% vs. 53, 63% of the respective primary diagnoses, p=0.04). The most frequently reported symptom associated with a clinical label of dysautonomia was exercise intolerance (78%), followed by fatigue, dizziness, gastrointestinal (GI) symptoms and palpitations (Table 2). Of individuals with a clinical label of dysautonomia, most (93%) reported more than one dysautonomia-related symptom with over half reporting more than five (Figure 1). Because exercise intolerance was so common in this cohort and hEDS/HSD patients often suffer with chronic musculoskeletal pain, we evaluated the self-reported explanation for limitation to exercise. Importantly, dysautonomia-related symptoms (50%) were implicated as frequently as pain (50%) in explaining patients’ inability to exercise at their baseline (Figure 2a). Of the dysautonomia symptoms restricting exercise, postural symptoms were the most common complaint (62%, Figure 2b). In total, 91 individuals (63.2%) met the study definition of dysautonomia and 53 (36.8%) did not and these designations were retained for downstream analyses. Clinical characteristics of individuals with and without dysautonomia were similar (Table 3).

Table 1.

Distribution of hypermobile Ehlers-Danlos syndrome and hypermobility spectrum disorder diagnoses.

Diagnosis	M/F/All (% of Total)
Hypermobile Ehlers-Danlos syndrome (hEDS) †	3/57/60 (41.7)
Generalized hypermobility spectrum disorder (G-HSD)	4/47/51 (35.4)
Localized hypermobility spectrum disorder (L-HSD)	1/16/17 (11.8)
Historical hypermobility spectrum disorder (H-HSD)	1/11/12 (8.3)
Asymptomatic generalized joint hypermobility (A-GJH)	0/4/4 (2.8)
Total	9/135/144 (100.0)

^†One Female to Male transgender patient.

Table 2.

Prevalence of dysautonomia-associated symptoms

Symptom	N (%)
Exercise intolerance	79 (78.2)
Fatigue	78 (77.2)
Dizziness	71 (70.3)
GI	62 (61.4)
Palpitations	54 (53.5)
Poor concentration	47 (46.5)
True/near syncope	38 (37.6)
Dyspnea	36 (35.6)
Chest pain/discomfort	34 (33.7)
Thermodysregulation	34 (33.7)
LE Edema	15 (14.9)
Hypotension	11 (10.9)
Urinary	9 (8.9)

Data expressed as total number and percentage of clinically labeled dysautonomia patients endorsing a symptom in each category; GI: gastrointestinal symptoms including nausea, vomiting/regurgitation, constipation, diarrhea, and/or abdominal pain; LE: lower extremity

Figure 1.

Number of dysautonomia symptom categories endorsed in hypermobile Ehlers-Danlos syndrome and hypermobility spectrum disorder patients with a clinical label of dysautonomia. Multiple symptomatology was the rule, with >50% of patients having more than five dysautonomia symptoms.

Figure 2.

Self-reported limitation to achieving pre-hypermobility Ehlers-Danlos syndrome or pre-hypermobility spectrum disorder onset/flare exercise level.

a) Barriers to exercise were either physical pain related to the underlying joint hypermobility-related disorder or dysautonomia symptoms.

b) Of those reporting dysautonomia symptoms as their major limitation to achieving pre-joint hypermobility-related disorder onset levels of exercise, most endorsed postural symptoms.

Table 3.

Characteristics of individuals with and without dysautonomia

	Female			Male
	Dysautonomia (N=87)	No dysautonomia (N=47)	P-value	Dysautonomia (N=4)	No dysautonomia (N=5)	P-value
Age (years)	33.5 (13–71)	31 (16–72)	0.7	28.5 (19–49)	22 (21–42)	0.4
Height (cm)	165.1 (148.5–183)	165.1 (149.9–180)	0.8	174.9 (173–187)	183.2 (175.3–187)	1.0
Weight (kg)	65.7 (38.1–100.4)	61.7 (42–106.1)	0.5	64.7 (54.2–75.7)	87.9 (58.6–106.5)	0.3
BSA (m 2 )	1.7 (1.3–2.3)	1.7 (1.4–2.3)	0.7	1.8 (1.6–1.9)	2.2 (1.7–2.3)	0.3
EF (%)	62.5 (50–76)	62.5 (54–73)	0.6	58.5 (57–60)	62.5 (58–65)	0.2

Data expressed as a median (range)

EF = ejection fraction, data available for 143/144 individuals

At the time of hEDS/HSD initial evaluation, >75% of subjects reported a sedentary lifestyle, meeting the self-reported physical activity level designation of “none” (N=49, 34%) or “light” (N=60, 42%); only 35 (24%) reported exercising at a “moderate” or “vigorous” level. In contrast, prior to the onset or significant worsening of musculoskeletal symptoms either in the setting of injury or otherwise (“hEDS/HSD onset/flare”), 60% of subjects exercised at a “moderate” (N=61, 42%) or “vigorous” (N=26, 18%) level; only 19 (13%) of patients performed no formal exercise prior to hEDS/HSD onset/flare. While 51 (56.0%) of dysautonomia patients were moderate/vigorous exercisers at baseline, this number dropped to 14 (15.4%) following hEDS/HSD onset/flare (Figure 3a). The difference was less dramatic among non-dysautonomia patients (Figure 3b), where 36 (67.9%) vs. 21 (39.6%) were moderate/vigorous exercisers at baseline compared to after hEDS/HSD onset/flare. Notably, the number of patients reporting no regular physical activity (“none”) at baseline was 15 (16.5%) among dysautonomia and 4 (7.5%) among non-dysautonomia patients. At the time of hEDS/HSD evaluation, 21% of patients reported engaging in a formal physical therapy (PT) program (focused on musculoskeletal symptoms); 3% of these reported significant improvement of arthralgia and improved exercise tolerance with PT, but none could perform at their baseline exercise level due to persistent dysautonomia-related symptoms.

Figure 3.

Exercise level categories of subjects with hypermobile Ehlers-Danlos syndrome and hypermobility spectrum disorder pre- and post- onset/flare of musculoskeletal symptoms. Subjects with a) dysautonomia displayed a more apparent shift towards “light” or “none” exercise levels following onset/flare of musculoskeletal symptoms compared to b) non-dysautonomia subjects.

LV chamber sizes were smaller overall for hEDS/HSD patients with dysautonomia compared to those without (Table 4). LV EDV/BSA was significantly lower in females with dysautonomia (57.8 ± 17.2 ml/m²) compared to those without (65.0 ± 15.1 ml/m²) (p = 0.03). For both males and females, LV internal diameter in diastole (LVIDd) and indexed LV mass trended lower in the dysautonomia vs. non-dysautonomia groups, but did not reach statistical significance.

Table 4.

Comparison of 2D echocardiography-derived left ventricular chamber size by gender and dysautonomia status.

	Female			Male
	Dysautonomia (N=87)	No dysautonomia (N=47)	P-value	Dysautonomia (N=4)	No dysautonomia (N=5)	P-value
LV EDV/BSA (ml/m 2 )	57.8 ± 17.2	65.0 ± 15.1	0.03	59.9 ± 10.4	73.3 ± 19.8	0.3
LVIDd (mm)	44.3 ± 6.0	46.7 ± 5.3	0.3	44.0 ± 2.6	48.2 ± 6.7	0.4
LV mass/BSA (g/m 2 )	65.8 ± 16.8	72.0 ± 18.5	0.2	74.4 ± 22.8	73.0 ± 13.6	0.9

Echocardiographic data expressed as mean ± s.d.

BSA: body surface area

LVEDV: left ventricular end systolic volume

LVIDd: left ventricular internal diastolic diameter

^†Echocardiographic data available for individuals with and without dysautonomia respectively for the following parameters: LVIDd, N=69 and N=36; LVEDV/BSA, N=60 and N=45; LV mass/BSA, N=67 and N=40.

When compared to measurements from a heart-healthy reference non-hypertensive control population (Cuspidi et al., 2012), LVIDd were smaller for dysautonomic females (44.3 vs. 46.0 mm, p=0.03) and males (44.0 vs. 50.0 mm, p=0.02) but not those without dysautonomia (46.7 vs. 46.0 mm for females, p=0.5; 48.2 vs. 50.0 mm for males, p=0.6). Mean indexed LV mass was significantly lower for dysautonomic females (65.8 vs. 72.3 g/m², p < 0.01) but not non-dysautonomic females or any males compared to the respective reference means (Supplemental Table 2).

Discussion

Hypermobility disorders are strikingly common but rarely diagnosed (Kumar & Lenert, 2017). A conservative estimate is that hEDS/HSD affects 3% of the population (Mulvey et al., 2013), making these disorders >100 times more prevalent than Marfan syndrome. Dysautonomia is considered a common comorbidity, but the prevalence in hEDS and HSDs has not been defined separately in the wake of contemporary nosology. Of the 144 patients in this study with one of these disorders, 63% had concomitant dysautonomia. Notably, the rate of dysautonomia was higher in females than males, although the small number of males in this study precludes meaningful statistical comparisons. As with JH disorders, symptom burden is higher in dysautonomic women, often those of child-bearing age (Garland, Raj, Black, Harris, & Robertson, 2007). Patients with JH-related disorders (Scheper et al., 2016) and those with POTS (Raj, 2013) report low health-related quality of life and are often unable to go to school, work or participate in recreational activities. As we found in this study and has been shown previously (Wallman et al., 2014), these disorders are often found in tandem, putting this population of largely young women at risk for significant morbidity and decreased quality of life over many years.

In the present study, dysautonomia was common among both hEDS and HSD cohorts, indicating a likely shared pathophysiologic mechanism, although the prevalence was significantly higher in hEDS patients. Still, our results show that dysautonomia cannot be used as a discriminator between these diagnoses. The dysautonomic symptomatology observed in our hEDS/HSD cohort was wide and can have a considerable impact on quality of life. Indeed, in our cohort, >90% of patients with dysautonomia had three or more related symptoms. Exercise intolerance was notably common among JH-related dysautonomia patients, occurring in >75%. While it is possible that exercise intolerance for some was due to physical limitations and pain from underlying hEDS/HSD, most individuals reported this as part of a collection of other classic dysautonomia symptoms, and only a handful cited exercise intolerance as their only symptom, fewer in dysautonomia vs. non-dysautonomia patients (0 vs. 5). This particular scenario inevitably invokes a “chicken or egg” causality quagmire: does exercise intolerance beget dysautonomia or vice versa? In a recent review of POTS where EDS and JH were recognized as frequently associated conditions, the most common demographic noted was young, previously active women with an identifiable event preceding symptom onset and resulting in bed rest or withdrawal of physical activity (Bryarly, Phillips, Fu, Vernino, & Levine, 2019). Determining the natural history of exercise intolerance in JH-related dysautonomia was outside the scope of the present study but warrants further examination since physical therapy exercises are the cornerstone of treatment for hEDS/HSD.

Dysautonomic mechanisms in POTS patients are postulated to relate to aerobic deconditioning as evidenced by the observation of lower physical performance due to lower stroke volumes during exercise (Fu et al., 2011; Shibata et al., 2012) and exercise training in POTS expands plasma volume resulting in increases in cardiac size and mass as well as improved orthostatic tolerance (George et al., 2016). These mechanisms have not been explored more broadly in dysautonomia patients, especially those with hEDS/HSD. However, our hEDS/HSD dysautonomia cohort displays many similarities to POTS patients, including overlap in symptoms, especially exercise intolerance and deconditioning (Fu et al., 2010). Furthermore, a significant proportion of patients in our cohort reporting high levels of physical activity and exercise at baseline endorsed a decline to none/light activity following a period of inactivity due to acute illness (e.g., motor vehicle accident, tendon tear, flare of gastrointestinal symptoms). In the POTS literature, a history of symptom exacerbation following a prolonged period of reduced activity is common (A. Hakim et al., 2017). This lends further credence to the hypothesis that aerobic deconditioning may underlie dysautonomic symptoms in hEDS/HSD. Considering that patients with hEDS/HSD suffer flares of musculoskeletal pain and frequent joint and soft-tissue injuries (Castori, 2016), such individuals would be at particularly high risk. Additionally, we found that patients without dysautonomia had overall higher exercise levels before and after the onset or flare of their hEDS/HSD compared to those with dysautonomia, with 92% vs. 84% reporting engaging in some formal physical activity and only 21% vs. 42% dropping down to no exercise thereafter. This raises the possibility that continued basal exercise throughout hEDS/HSD flares may be protective against the development of dysautonomic symptomatology. Further prospective case-control studies are needed to investigate this prospect.

The mechanisms underlying autonomic dysfunction in hEDS/HSD remain unclear and this study was not designed to define them, but are suggested to include low blood pressure, increased peripheral venous dilation and blood pooling, low circulating blood volume, elevated circulating catecholamines and excess systemic levels of histamine (A. Hakim et al., 2017). There is evidence supporting some of these mechanisms in hEDS such as increased aortic wall compliance (Handler, Child, Light, & Dorrance, 1985), abnormal connective tissue in dependent blood vessels causing excessive distention in response to ordinary hydrostatic pressures (Rowe et al., 1999), neuropathy, connective tissue laxity, alpha and beta adrenergic hyper-responsiveness (De Wandele et al., 2014; Mathias et al., 2011), as well as potentially pathogenic adrenergic, muscarinic and other neural autoantibodies (A. Hakim et al., 2017). Proposed mechanisms for POTS specifically include small cardiac chamber size and low plasma volume (Fu et al., 2010), and considering the symptom overlap, it is reasonable to consider that these features may also be part of the hEDS/HSD dysautonomia phenotype. Our findings of smaller LV volumes in hEDS/HSD patients with vs. without dysautonomia and overall LV chamber sizes lower than reference normal in patients with dysautonomia but not those without supports this theory and warrants further study using a standardized prospective design.

Limitations

This retrospective study has several limitations inherent to the design. Both hEDS/HSD and dysautonomia are clinical diagnoses of exclusion that are challenging to make and no tools are currently available to objectively define them. Formal autonomic testing was not performed for many of the subjects, although such testing is insufficient for diagnosing or ruling out dysautonomia at large. Echocardiograms were performed in the clinical setting and not standardized through a core laboratory for this retrospective study. Finally, since males with hEDS and HSDs tend to have fewer symptoms (Castori, Camerota, Celletti, Grammatico, & Padua, 2010) they are seen less often in the clinical setting. Small numbers of males in this study limits the applicability of these findings to that population.

Conclusion

In summary, patients with either hEDS/HSD or dysautonomia are at heightened risk for substantial morbidity and impaired quality of life, but carrying both diagnoses imposes an even greater symptomatic burden and disproportionately affects young women. hEDS and HSD lead to exercise intolerance and avoidance, likely further potentiating symptoms and sparking an unfortunate cycle with sedentary lifestyle as a cause and consequence. Cardiac atrophy is a feature of this phenomenon and aerobic reconditioning to more favorably remodel the cardiac geometry may be a promising intervention that deserves further investigation.